There are many human physiological characteristics which can be used to provide personnel identification for security purposes, such as fingerprint, retina, iris, DNA, or even face features. For all the devices that are capable of distinguishing some physiological characteristics of one person from others', a fingerprint reader has the lowest cost and complexity, while the identification results are generally pretty good. In addition, the size of data required to store the minutiae of one fingerprint is small (ranging from 120 bytes to 2K bytes). This makes fingerprint identification devices widely accepted in many applications.
There are also many types of sensing techniques for capturing fingerprints. The popular ones are the optical type and capacitive type. Optical fingerprint sensing modules utilize reflected light intensity from the surface of a finger to tell where the ridges and valleys are on the contact portion of the finger. The advantage of the optical technique is reliability and low cost. However, due to the size of the embedded optical lens, the form factor of an optical fingerprint sensing module cannot be kept small. It is difficult for the optical type sensor to be embedded in portable devices. The capacitive-type fingerprint identification modules, on the other hand, are made out of silicon chips and can be made very compact. In some cases, when a fingerprint image can be fetched by slide scanning, the fingerprint sensor can be even thin and slim, too. The small form factor of capacitive-type fingerprint identification module makes it suitable for portable applications such as access control badges, bank cards, cellular phones, tablet computers, USB dongles, etc.
Capacitive-type fingerprint sensors are based on the principle that the capacitance of two parallel conductive plates is inversely proportional to the distance between them. A capacitive fingerprint sensor consists of an array of sensing units. Each sensing unit contains a sensing plate. By using the sensing plate as one plate of the two-plated capacitor and a dermal tissue as another plate, ridges and valleys of a fingerprint can be located by measuring the different capacitances. There are many prior arts related to the capacitive-type fingerprint identification module. Most of them have been applied to manufacture fingerprint sensors. However, there are also many problems pending for solutions. One of them is the accuracy of the sensing elements.
Due to the high-density nature, the popular capacitive fingerprint sensors are mainly manufactured with semiconductor processes. The precision of the sensing elements is affected by many factors inherited in the manufacturing technology, such as density of chemical impurities, alignment of photomasks, equipment control, etc., whose uncertainty or variation will be reflected in the different behavior between devices, or even a fixed pattern noise seen in the captured fingerprint images of the same device. In order to achieve the best performance of personal identification, it is desirable to improve the quality of the captured fingerprint image by reducing the noise pattern. A common practice to eliminate fixed pattern noise is to calibrate the device before use. The calibration data can be calculated and stored as part of the manufacturing process, or right before the device is used. However, in either case, a certain amount of memory storage space must be set aside for the calibration data, and this storage space will increase the system cost. Therefore, an innovative pixel sensing element, a capacitive fingerprint sensor made by the pixel sensing elements and a method for running the pixel sensing element are desirable.